The purpose of this project is to determine the reliability and validity of an instrument designed to assess functional ability of patients with Alzheimer's Disease (AD) enrolled in clinical treatment trials. At present functional scales measuring activities of daily living (ADLs) are often used as secondary but rarely as primary outcome measures either in trials of antidementia drugs or in trials of drugs for the treatment of ancillary symptoms such as agitation. Existing instruments are limited first because there is no consensus that they measure all of the major functional limitations found in AD patients and more importantly, that no existing instrument has been shown to be suitable for assessing functional change over a broad spectrum of dementia severity. Previously we have conducted reliability and longitudinal validity studies of the two most widely used instruments for assessing activities of daily living (ADLs) in AD patients, the Physical and Self-Maintenance Scale (PSMS) and Instrumental Activities of Daily Living Scale (IADLS) of Lawton and Brodie (1969). These studies have determined that many but not all of the items on these two scales have adequate reliability and longitudinal validity. We propose to modify some of the PSMS items to enhance reliability and to modify or replace several IADLS items so that they are less gender and location specific. Then we propose to assess the inter-rater, test-retest and alternate source reliability of each item; those with poor reliability will be eliminated or modified. Then a group of at least 50 AD patients with a wide range of baseline severity will be evaluated every six months for a minimum of 12 months. The ability of each item to measure longitudinal change for each level of dementia severity will be determined, and the reliability and sensitivity of the scale will be compared with other clinician rating scales used in clinical trials. The ultimate aim of the project is to produce an instrument for measuring activities of daily living (ADLs) in AD patients that can be used to evaluate the efficacy of drug treatments or other interventions.